About neural network : newbie question

Question

Parkz Leingodz 2013-12-18

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/110122-about-neural-network-newbie-question

评论： Parkz Leingodz 2013-12-20

I am a new to matlab. I just covert my image to matrix and want to train it. here is my coding:

P= [];
for i =1:15
str=strcat('E:Skripsi\crop2\image',int2str(i),'.bmp');
eval('img=imread(str);');
 img=reshape(img,[],1);
    P= [P img];
end
T=[1 1 1 1 1 1 1 1 1 1 0 0 0 0 0];// if true image i make it 1 , negative image make it  0
p = num2cell(P,1);
t = num2cell(T,1);
net = newff(minmax(p), [3 1], {'logsig' 'purelin'});
net.trainParam.epochs = 10;
net = train(net,p,t);
Y=sim(net,p)

these is what i got when i compile it

 [0.9794]    [0.9974]    [1.2784]    [1.0014]    [1.0058]    [0.9904]    [0.7293]    [1.0001]    [1.0106]
Columns 10 through 15
    [1.0002]    [0.0618]    [3.3131e-04]    [9.9559e-04]    [0.0134]    [0.0178]

then my problem is what should i do to make the result success i train in percentage with error below 1% (what i want: example i train 10 true image to 1 and one result of my train is 0.798 then accurate =90%) thanks for the help.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Greg Heath 2013-12-19

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/110122-about-neural-network-newbie-question#answer_118808

There is no reason to use cells or only train for 10 epochs.

Use round to convert the output to {0,1}

Thank you for formally accepting my answer

Greg

5 个评论
显示 3更早的评论隐藏 3更早的评论

Parkz Leingodz 2013-12-19

在 MATLAB Online 中打开

here is my non cellcode at first :

P= [];
for i =1:15
str=strcat('E:Skripsi\crop2\image',int2str(i),'.bmp');
eval('img=imread(str);');
 img=reshape(img,[],1);
    P= [P img];% convert all the image to one matrix
end
T=[1 1 1 1 1 1 1 1 1 1 0 0 0 0 0]; 
net = newff(minmax(P), [3 1], {'logsig' 'purelin'});
net.trainParam.epochs = 10;
net = train(net,P,T);
Y=sim(net,P)

and if i compile it , here is the error :

Error using newff>new_5p0 (line 257)
Inputs are not a matrix or cell array with a single matrix.
Error in newff>create_network (line 128)
  net = new_5p0(varargin{:});
Error in newff (line 102)
  out1 = create_network(varargin{:});
Error in f (line 12)
net = newff(minmax(P), [3 1], {'logsig' 'purelin'});

actually i don't know how to fix this error and where's the error, so i just try to change P to cell and it works.

but what u mention earlier is true, i have to change cell2mat when i want get {0,1}. so do you have an advice to fix my non cell code?

i don't know how to answer what dimension P is. i just want to convert image to one matrix and train it. thanks

Greg Heath 2013-12-19

whos P

size(P)

Parkz Leingodz 2013-12-20

size P is 16x21 pixels.

P is a matrix that save whole matrix value of img.

请先登录，再进行评论。

About neural network : newbie question

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

5 个评论
显示 3更早的评论隐藏 3更早的评论

更多回答（0 个）

另请参阅

类别

标签

产品

Community Treasure Hunt

About neural network : newbie question

0 个评论 显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

5 个评论 显示 3更早的评论隐藏 3更早的评论

更多回答（0 个）

另请参阅

类别

标签

产品

Community Treasure Hunt

WeChat

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

5 个评论
显示 3更早的评论隐藏 3更早的评论